1133

There are on average 8 infant deaths a year in South Australia that might be avoided by ECMO-ie,1/3000 births, which is lower than Bardett and co-workers’ prediction, mainly because of the lower incidence of meconium aspiration syndrome. Nevertheless, such an incidence would support the introduction of ECMO as a clinical service. Redhill Hospital, Redhill, Surrey RH1 6LA, UK

1. Bartlett RH,

I. G. LEWIS

Gazzaniga AB, Toomasian JM. Extracorporeal membrane oxygenation respiratory failure: 100 cases. Ann Surg 1986; 204: 23.

m neonatal

Tolerance and the fetal

graft

Characteristic nuchal oedema with trisomy 21.

(arrow)

at 11.2 weeks in a fetus

.

SiR,—Your Sept 1 editorial discusses the paradox of the fetal allograft and in particular the possible role of the HLA-G antigen on the cytotrophoblast. You conclude that a special nonimmunogenic form of the major histocompatibility complex (MHC) class I molecule expressed at the fetomaternal interface prevented the baby being rejected as foreign by his mother. We question that such a molecule does not elicit a maternal immune response. Many workers have shown that antibodies to paternal antigens can be detected during pregnancy.1-3 We have demonstrated a non-cytotoxic IgG antibody response to paternal lymphocytes in sera from first trimester primigravidae4 and in placental eluates from these women at term.s The antibodies were directed to HLA-linked antigens but not to recognised class I (HLA-A, HLA-B, and HLA-C) or class II (HLA-DR) specificitiesAlthough the target for these antibodies has not been further defined, our data suggested that it may be an antigen similar to the special MHC molecule considered by you. Therefore such a molecule might be responsible for eliciting the antibody response we described. These antibodies are non-cytotoxic and might not be harmful to the survival of the fetal allograft.

Department of Renal Medicine, City Hospital, Nottingham NG5 1PB, UK

ANDREW INNES

Department of Medicine and Therapeutics, University of Aberdeen, Aberdeen

DAVID POWER GRAEME CATTO

CHARLES CUNNINGHAM

H, Perry WM, Rocklin RE. Suppression of maternal responsiveness to paternal antigens by material plasma J Immunol 1975; 114: 525-28. 2. Rocklin RE, Kitzmiller JL, Garovoy MR. Maternal-fetal relation: II, further characterisation of an immunological blocking factor that develops during pregnancy, Clin Immunol Immunopathol 1982, 22: 305-15 3 Power DA, Catto GRD, Mason RJ, MacLeod AM, Stewart GM, Shewan WG. The fetus as an allograft: evidence for protective antibodies to HLA-linked paternal antigens. Lancet 1983; ii: 701-04. 4. Innes A, Cunningham C, Power DA, Catto GRD. Fetus and an allograft: non-cytotoxic maternal antibodies to HLA-linked paternal antigens. Am J Reprod 1. Pence

Immunol 1989; 19: 146-50. A, Stewart GM, Thomson MAR, Cunningham C, Catto GRD. Human placenta an antibody sponge? Am J Reprod Immunol 1988; 17: 57-60.

5 Innes

Nuchal fluid accumulation in trisomy-21 detected by vaginosonography in first trimester SIR,-Fetal anatomy in the first trimester can be seen with a high degree of accuracy by vaginosonography.1,2 During the past three years we have examined with sonography 7 fetuses with trisomy 21 diagnosed by karyotyping in our first-trimester chorionic villus sampling programme. In all 7 fetuses we demonstrated accumulation of subcutaneous fluid in various amounts in the nuchal region (figure). However, of 105 control fetuses with normal karyotype matched for gestational age (11 and 12 weeks) excessive nuchal fluid was detected in only 1. In the fetuses with normal karyotype the nuchal thickness was less than 2-5 mm, whereas in the affected fetuses this measured between 3 and 7 mm. Redundant nuchal folds are among the characteristic features of second-trimester fetuses with trisomy 21; however, this sign alone is

Sagittal plane. of little value in screening for Down syndrome. Spontaneous resolution of first-trimester nuchal oedema later in pregnancy has also been described/3 and greater attention should be paid to this sign in the first trimester. Aneuploidy cannot be assumed to be present in every fetus with this appearance but it is highly probable that fetal nuchal oedema detectable by vaginosonography in the first trimester is a manifestation of trisomy 211,2 or other aneuploid syndromesand could be an indication for chromosomal studies. Vaginosonography between weeks 9 and 12 of gestation might result in a higher pick-up rate of abnormalities than other methods, and this could be a promising approach in screening for trisomy 21 in patients at risk and in young pregnant women, since the procedure is non-invasive and therefore less risky than other methods. Department of Obstetrics and Gynaecology, Medical University A Szent-Gyorgyi, 6701 Szeged, Hungary

J. SZABÓ J. GELLEN

1. Rottem

S, Bronshtem M, Thaler I, Brandes JM. First trimester transvaginal sonographic diagnosis of fetal anomalies. Lancet 1989; i: 444-45. 2. Bronshtein M, Rotten S, Yoffe N, Blumenfeld Z. First-trimester and early second trimester diagnosis of nuchal cystic hygroma by transvaginal sonography: diverse prognosis of septated from nonseptated lesion. Am J Obstet Gynecol 1989; 161: 78-82. 3. Rodis JF, Vintziloes AM,

Campbell WA, Nochimson DJ. Spontaneous resolution of fetal cystic hygroma m Down’s syndrome. Obstet Gynecol 1988; 71: 976-77. 4. Szabó J, Gellén J, Szemere G. Nonimmune hydrops in trisomy 18: diagnosis by chorionic villus sampling and vaginosonography in the first trimester. Br J Obstet Gynaecol (in press).

Bone density screening for osteoporosis SiR,—We are afraid that Dr van Hemert’s reply (Sept 29, p 818) to views (Aug 25, p 502) will add further confusion to what should straightforward approach to bone density screening for osteoporosis. Osteoporosis is a condition that is identified by the presence of a low bone mineral density (BMD), a consequence of the decline in bone mass per unit volume. Whether that disease expresses itself as a fracture in the future depends on several other factors such as a continuing adverse lifestyle and absence of sex hormone replacement therapy (contributing to further bone loss), a fall, or a lack of muscle and fat which may protect against the consequences of a fall. Van Hemert confuses the diagnosis of the disease by BMD scanning with risk evaluation for osteoporosis. We are agreed that clinical risk factors have poor diagnostic value but to use, as he does, metacarpal cortical bone loss measured by a magnifying glass from X-ray films of the hands as a predictive test is mistaken.1,2 Trabecular bone loss is the major feature of postmenopausal osteoporosis due to oestrogen deficiency and is the main reason for the greatly increased prevalence in women. Dual photon bone densitometry screening identifies this trabecular loss in the spine or hip, the critical areas of fracture. His work did not test this predictive performance. The important epidemiological question is not the overlap in BMD between fracture and non-fracture groups but

our

be

a

1134

whether fracture rates in postmenopausal women who have preserved their BMD are equivalent to those in young normal women. Epidemiological evidence suggests that this is likely to be SO.3 To assert that osteoporosis "seldom leads to death" is to ignore UK Department of Health figures for 1985: there were 44 000 fractures of the hip and almost 9000 of those patients died and 22 000 became invalids, dependent on relatives or-the State.4 Van Hemert himself really makes the point for bone density screening when he emphasises that all menopausal women have bone loss which is likely to be very difficult to restore later. The point of screening is to identify the problem early, when sex hormone replacement will prevent further bone loss for as long as it is taken.5 Where he is mistaken is in assuming that all women lose bone at 1 % per year, taking 30 years to lose 30%-losses which he says cannot be restored. He seems to feel that this is the critical level of loss. In our osteoporosis and menopause clinic we have seen patients who have lost up to 18% of bone in 6 months, while others on treatment have increased BMD by up to 15% in one year, thereby significantly diminishing their risk of fracture. We agree that a 1 % loss of bone will have little influence on the risk of fracture, but 10% will, by a factor of 2-3-fold.6 Van Hemert does not say how he would discover the "rapid bone losers" without dual photon bone density screening and follow-up scans. We replied in detail previously to the difficulty of offering oestrogen therapy to all menopausal women, concentrating instead on those with osteoporosis as an identifiable group at risk. In such patients the logic of fracture prevention is more likely to prevail over personal, although not necessarily well informed, preference to do nothing and hope for the best. The continuing great reluctance of doctors and patients to consider osteoporosis screening and sex hormone replacement will most effectively be answered by the growing realisation of the prevalence of this condition in women, with rising mortality and morbidity. We hope that this will not take 30 years nor be considered a matter unworthy of attention by a caring profession, as van Hemert’s words might imply: "osteoporosis, though a common disease, occurs almost exclusively among old to very old people". Endocrine and Dermatology Centre, 140 Harley Street, London W1 N 1AH, UK

M. I. ANDERSSON W. PERRY C. H. MORTIMER

1. van Hemert AM, Vandenbroucke JP, Birkenhager JC, Valkenburg HA. Prediction of osteoporotic fractures in the general population by a fracture nsk score: a nine-year follow-up among middle-aged women. Am J Epidemiol 1990; 132: 123-35. 2. van Hemert AM, Vandenbroucke JP, Hofman A, Valkenburg HA. Metacarpal bone loss in middle-aged women: "horse racing" in a 9-year population based follow-up study. J Clin Epidemiol 1990; 43: 579-88. 3. Kiel DP, Felson DT, Anderson JJ, Wilson PWF, Moskowitz MA. Hip fracture and the use of estrogens in postmenopausal women: the Framingham Study. N Engl J Med 1987; 317: 1169-74. 4. Consensus Conference on Osteoporosis. Organised by National Osteoporosis Society (Oct 6, 1987). London: Medicine Group, 1987. 5. Lindsay R, Aitken JM, Anderson JB, Hart DM, MacDonald EB, Clarke AC. Long-term prevention of postmenopausal osteoporosis by oestrogen. Lancet 1976; i: 1038-41. 6. Melton LJ, Wahner HW, Richelson LS, O’Fallon WM, Riggs BL. Osteoporosis and the risk of hip fracture. Am J Epidemiol 1986; 124: 254-61.

Paroxysmal dystonic reflex choreoathetosis after minor closed head injury

department where he was examined and released with only mild neck stiffness. 3 days later, he noted intermittent, brief, paroxysmal episodes of dystonia of his right neck, with an urge to turn his head to the right. 5 days after the accident, while sitting, he had the sudden onset of intense right shoulder pain followed by paroxysmal dystonic movements of his right face, neck, shoulder, arm, and leg, lasting 5-10 s. He could briefly suppress the movements with intense concentration. They occurred at rest and did not increase with movement. His right shoulder pain subsided after 15 minutes and was replaced by a right hemisensory dysaesthesia to pinprick and light touch. His medical history was unremarkable, with no history of cocaine or neuroleptic drug use. He had no family history of movement or degenerative neurological disorders. He had a normal birth and developmental history. After the accident he had taken cyclobenzaprine 10 mg three times a day and diflunisal 500 mg twice daily, but had stopped these after 3 days. Examination revealed normal cranial nerves, motor system, reflexes, and cerebellar function with no evidence of trauma or neurocutaneous lesions. He had decreased pinprick and light touch sensation over his entire right face and body. Any new light touch on his right face or shoulder triggered paroxysmal dystonic reactions. He accommodated to constant touch after one or two dystonic spells and could wear clothes. His gait was normal when concentrating. Otherwise, he had severe paroxysmal dystonic movements of his right body. Laboratory tests were normal though a urine drug screen revealed cannabinoid byproducts. Cranial and cervical spine magnetic resonance imaging scans and an electroencephalogram during episodes of dystonia were normal. The frequency of his dystonic spells decreased over several weeks in hospital while taking diazepam tablets 10 mg thrice daily. After a week the right hemisensory deficit resolved. At 2 month follow-up he was symptom-free on no medication. Dystonic movements can be produced by damage to the basal ganglia or through damage to peripheral and cervical regions 4,s Acquired dystonias have been associated with perinatal brain injury, carbon monoxide anoxia, encephalitis, manganese toxicity, drugs, strokes, arteriovenous malformations, brain tumours, and head trauma.6 Most reported cases of dystonia are fixed and non-fluctuating.4,7 Previous descriptions of head injury causing paroxysmal dyskinesias report the dyskinesias as lasting minutes to hours, much longer than the spells in our patient. His movements were atypical of acquired secondary dystonias being sudden in onset and shortlasting ; he was also normal between spells. The dyskinesia was not stimulated by movement and was thus not kinesogenic. Nor does the patient seem to have had a primary dystonia that was unmasked by the accident (eg, paroxysmal dystonic nonkinesogenic choreoathetosis or exertional paroxysmal choreoathetosis6). He may have had a form of reflex epilepsy, a term used to describe seizures with specific stimulations (photic, startle, music, reading, writing)89 but this is unlikely because of the normal EEG during the spells. A conversion disorder is unlikely since clinical examinations were consistent over time and with different physicians, and his responses could not be changed by bedside techniques. This is, to our knowledge, the first reported case of paroxysmal dystonic reflex choreoathetosis after minor head injury. University Department of Clinical Neurology, Institute of Neurology, London WC1N 3BG, UK

J. B. PICKETT

SIR,-Dystonic movements have been reported after closed head injury with damage to basal ganglia structures" but all previous followed major

trauma with loss of consciousness, and the lasted for several minutes at least. We have treated a patient with unilateral paroxysmal dystonic choreoathetosis which could be triggered by tactile stimulation of the right shoulder and body. The dystonic attacks lasted only 5-30 seconds, but happened about thirty times a day, and followed minor closed head injury without loss of consciousness. A 33-year-old man was in good health when a heavy lorry struck his car from behind. He suffered minor cervical sprain without striking his head and was not unconscious. He drove to a casualty cases

dystonic

movements

MARK S. GEORGE

Medical University of South Carolina, Charlston, South Carolina, USA 1.

2.

H. KOHLI M. A. ALLISON PAUL PRITCHARD

Pettigrew LC, Jankovic J. Hemidystonia: a report of 22 patients and a review of the literature. J Neurol Neurosurg Psychiatry 1985; 48: 650-57. Robin JJ. Paroxysmal choreoathetosis following head injury. Am Neurol 1977; 2:

447-48. K, Cohen JA. Post traumatic torticollis. Neurology 1989; 39: 1642-43. 4. Brett EM, Sheehy MP. Progressive hemidystonia due to focal basal ganglia lesions after mild head trauma. J Neurol Neurosurg Psychiatry 1981; 44: 460. 5. Marsden CD, Obeso JA, Zarranz JJ, Lang AE. The anatomical basis of symptomatic hemidystonia. Brain 1985; 108: 463-83 3. Isaac